Microsoft has confirmed that its August 2025 Patch Tuesday updates broke the Reset this PC recovery feature on Windows 11 and Windows 10, and an out-of-band emergency fix is being prepared. The breakdown affects millions of devices that have installed the mandatory security updates KB5063878 (Windows 11 24H2), KB5063875 (Windows 11 23H2/22H2), and KB5063709 (Windows 10). Within days of deployment, a separate wave of user reports began surfacing that the same August update for Windows 11 24H2 can destabilize certain NVMe SSDs, causing drives to vanish or return corrupted data under heavy write loads. The dual failures strike at the core of system reliability, leaving users with broken recovery paths and potential storage carnage.

The recovery regression is official. Microsoft published an advisory stating that after installing the August updates, “attempts to reset or recover the device might fail.” The flaw renders Reset this PC, the Fix problems using Windows Update tool, and some remote wipe operations unusable. When invoked, the operations abort and roll back silently, offering no on-screen explanation. For enterprise IT teams and consumers alike, this defeats a critical last resort — the ability to reinitialize a damaged system or prepare a machine for hand-off. The company has committed to an out-of-band update that will land “in the coming days,” though no exact date has been given.

Simultaneously, a storage instability investigation is unfolding. Users on Reddit, technical forums, and testing labs have documented scenarios where large, sustained file transfers — commonly exceeding 50 GB of contiguous writes — trigger NVMe drives to disappear from Windows after installing KB5063878. Some drives reappear after a reboot; others require a complete power cycle. A handful of reports describe unrecoverable data corruption. SSD controller vendor Phison has acknowledged it is working with Microsoft and other partners to diagnose the issue. Testing by independent outlets like Tom’s Hardware and BornCity has reproduced the problem under controlled conditions, though Microsoft has not yet formally confirmed a universal storage defect. The working hypothesis points to an interaction between the updated Windows I/O stack and specific SSD firmware behaviors, especially on DRAM-less drives nearing capacity.

These problems landed during a routine security patch cycle. On August 12, 2025, Microsoft shipped the cumulative updates as part of its Patch Tuesday cadence. The packages were mandatory, closing several actively exploited vulnerabilities. For Windows 11 24H2, KB5063878 brought the build to 26100.4946. KB5063875 for Windows 11 23H2 and 22H2, and KB5063709 for Windows 10 followed suit. Initially, the official KB articles stated “Microsoft is not currently aware of any issues.” That language remained in place even as community reports and independent media exposed the problems, creating a messaging gap that frustrated users.

Microsoft’s communication has been a mixed bag. On one hand, the company rapidly acknowledged the recovery regression through its release health channels and promised an out-of-band fix. On the other, the primary KB pages for the affected updates still display the “no known issues” disclaimer, and the storage instability remains in an investigative limbo. This fragmented approach forces users to rely on tech news sites and forums for actionable intelligence. Enterprise customers can lean on Known Issue Rollback (KIR), which Microsoft offers as a surgical, auditable mitigation. Consumers, however, often stumble into broken recovery functions unaware.

The practical fallout is severe. A failed Reset this PC can leave a device in an unrecoverable state, especially if the user has no recent backup. IT administrators overseeing fleets managed via WSUS or SCCM face an additional headache: some deployments failed with error 0x80240069, a separate installation glitch that further gums up patch distribution. Meanwhile, the SSD quirk threatens data integrity for anyone who regularly pushes large files — video editors, developers juggling virtual machines, gamers downloading massive titles. In the worst case, a disappearing drive during a critical write operation could corrupt the file system.

For users and admins, the immediate guidance is clear. Do not use Reset this PC or the Windows Update troubleshooter on any device that has received the August update. Hold off on large file operations if you are on Windows 11 24H2 with an NVMe drive, particularly if the drive is more than 60% full or uses a Phison controller. Make a fresh, verified backup of all important data before attempting any recovery or major disk activity. Enterprises should pause automatic approval of the August cumulative, limit deployment to a pilot ring, and rely on manual installs from the Microsoft Update Catalog for critical hosts that cannot delay security patches.

To check if you are vulnerable, open Settings > Windows Update > Update history and look for the August 12, 2025 KB numbers. On Windows 11 24H2, build 26100.4946 confirms the KB5063878 installation. If a reset has already failed, stop retrying. Repeated attempts can deepen the corruption. Instead, use pre-update bootable rescue media or backup images to recover. Without a backup, professional data recovery services may be the only option.

Technically, how did we get here? The recovery regression likely stems from a servicing stack metadata mishap in the cumulative update. Recovery operations rely on orchestrated mounting of image files and component replacement, and a change in the LCU or SSU component could break that delicate sequence. Similar regressions in the past have produced installation errors in WSUS/SCCM contexts, so the pattern is familiar. The SSD instability is harder to pin down. Modern NVMe drives depend on a tight dance between Windows’ storage driver, the NVMe command queuing, and the drive’s firmware. A tweak to I/O scheduling or caching logic — even a small one — can expose latent bugs in controller firmware. Phison’s involvement suggests the problem may be concentrated in certain controller generations, but other reports hint at broader impact.

Microsoft’s incident response has strengths. The KIR mechanism allows enterprises to surgically disable the offending change without uninstalling the entire security update, preserving protection while awaiting a fix. The public acknowledgment of the recovery bug, albeit delayed, is a step toward transparency. Yet there are glaring weaknesses: the initial “no known issues” labeling on KB pages eroded trust, and the absence of a visible warning inside the Reset UI leaves unsuspecting users dead-ended. Repeated quality snafus in the monthly security rollups raise questions about pre-release testing, particularly around hardware diversity and enterprise management topologies.

What should you expect next? Microsoft will push out an out-of-band update for the recovery regression via Windows Update and the Microsoft Update Catalog. It may arrive as a small servicing stack update or a corrected cumulative. For the storage problem, SSD vendors will likely issue firmware updates if the root cause is confirmed on their side. Microsoft may release a partial mitigation through Windows Update or a driver update. In the meantime, the safest course is defense in depth: backup rigorously, stage rollouts carefully, and monitor both the Windows Release Health dashboard and vendor support pages.

This August incident is a sobering reminder that even critical security patches can destabilize the very systems they are meant to protect. The Windows ecosystem’s scale and hardware diversity guarantee that edge cases will slip through, but the pattern of repeated regressions demands better validation. For now, the burden falls on users and IT teams to stay vigilant, move cautiously, and keep their backups fresh. Microsoft’s promised emergency fix will restore the Reset function, but the SSD question remains an open wound — one that could inflict real data loss before a permanent cure arrives.